In these days of virtual server environments and data warehouses, keeping a data center going calls for a variety of high-availability solutions. Protecting against natural disasters — earthquakes, excessive cold or heat, flooding, hurricanes, lightning strikes, wildfires, etc. — is just one consideration. Business continuity and network uptime can be undermined by more mundane causes, such as disk failures, faulty components, and power outages. Here are five things to keep in mind on the subject of high availability.
- Have a disaster recovery plan in place before the disaster. Natural disasters are big news. While many disasters are rare occurrences, IT administrators need to be ready for what may happen in their locales, whether it’s big snowstorms, high water, or wind damage. Disaster recovery plans should be carefully considered and put in place. Getting the data out of the data center is a primary concern, and calculating the cost of downtime is another. Once you have a disaster recovery plan, don’t let it sit on a shelf and gather dust; continuously update the plan and expand it as needed.
- Make use of database mirroring and failover clustering. Database mirroring offers a duplicate copy of the database and can run on standard servers. Failover clustering, whether in automatic or manual modes, provides transparent client redirect.
- Log shipping can be an alternative to database mirroring — or a supplemental tool. Among other features, log shipping supports multiple secondary databases on multiple servers for a single primary database.
- Don’t forget about replication! Data replication allows for more than one redundant copy of the database.
- Always be thinking about improving high availability. As data centers are always changing and being updated, so too your high-availability solutions must be continuously changed, maintained, and updated.